Newbie Pro Tools latency question - need your help!

[Johnkenn]

Hey guys - just bought an Mbox 2 so I could add it to the old arsenal. I'm basically wanting to use the Mbox as a dongle, as I'm going from my pre - compressor - Rosetta 200 then spdif out into the Mbox. Then in order to use my Rosetta DA's, I'm going spdif out of the M?box into the Rosetta. All is fine and dandy except for the horrendous latency. Is there any way to overcome this with my setup? I can't belive PT's doesn't have direct monitoring? Maybe this is a lame question that has been answered before, but if someone can help me out, it would be MUCH appreciated. I've been avoiding PT's, but if I can get this fixed, I'd be pretty happy. BTW - yes, I've tried reducing my buffers to 128 - but there's still latency and that doesn't seem to be a very practical workaround.

Thanks -
John

Mac Pro
PTLE
3 ghz RAM

[Raw-Tracks]

Have you tried checking "Low Latency Monitoring" under the "Options" menu? That should do the trick.

[Ronster]

i'm not sure about the mbox 2, but i've used the first mbox and i always had horrible latency issues with it. after doing some research on google and on the digi forum i found out 2 things:

1) mbox has a latency of something like 164 samples or something like that...after reading up on it i found out that i had to nudge each track back 164 samples after recording it

2) the best way to really avoid horrible latency that occurs in the monitoring was for me to mute the track that was in record mode while i was recording...i know it sounds strange but it worked.

don't know if any of this applies to the mbox2 but i just thought i'd throw these things out there in case it might help you...hopefully someone can chime in and confirm this. let me know if any of this helps with the mbox2!

[Johnkenn]

Thanks, guys...I forgot to mention that the phone jack on the Mbox is disabled with me clocking with the Rosetta. I'm using the headphone pot on my Big Knob to monitor. I'll try the Low latency option under Options...

[mookie]

i'm in a similar situation. unfortunately, low latency is only available with the digi 001 or 002. the mboxes have "zero latency", but only via the headphone jack from the analogue inputs.
in the future, i'm going to try to split my preamp signal and run it into the analogue ins just to monitor. in the meantime (while i wait for the splitter), i'm just running at 128 samples... it's not so bad.

[Ronster]

have you guys tried to nudge each track back 164 samples after recording it? see if that's the key?

[Johnkenn]

I'm getting latency monitoring vocals when I sing...Sounds like I've got a delay on the track. I see why they call it zero latency for the headphones because it monitors the signal before it goes into PT. The problem is, when you bypass mbox's internal clock, it disables the headphone jack. (Because it's just passing digital info)
I was excited about getting PTs, but this is just unmanageable...I guess they only have ADC in the HD rigs to encourage you to buy the HD rig...then you have to buy one piece of Digi harware with it...I thought I would be able to use the mbox as a digital go between - basically use my Chandler and Vintech and Distressor and Apogee and then pass through the mbox like a $500 dongle. But if I can't find a workaround, I just don't see how this could be usable.
BTW - would this path solve everything? pre to compressor to Apogee spdif to a mixer with headphone jack then Spdif to the mbox?

[Johnkenn]

Another question -
What is the low latency monitoring on 002r and 002? If I upgraded, would I be dealing with this latency?
In Cubase, I can either monitor from Cubase (and hear the latency) or deselect the button and monitor directly what's coming in. Pro Tools doesn't have that ability? Are you kidding me? Why would anyone deal with LE if that's the case?

[paddo]

Man save yourself the trouble LE sucks the only PT worth dealing with is big bucks HD. For whatever reason,PT don't even have a decent core audio driver.And whilst i even considered working round the no ADC issue,Mbox records out of time 'are they kidding' how bad is that.




Paddo

[NickH]

Originally Posted by Johnkenn

Another question -
What is the low latency monitoring on 002r and 002? If I upgraded, would I be dealing with this latency?

This link will answer your questions on the latency figures:-

002/002R Latency & Low Latency Information

Low Latency Mode only works through the 002R’s analogue outputs 1 & 2…

So when tracking monitor through the 002R & when mixing monitor through the Apogee…

Cheers

N

[NickH]

Originally Posted by Ronster

have you guys tried to nudge each track back 164 samples after recording it? see if that's the key?

Pro Tools LE automatically compensates for any hardware delay after recording & does its own automatic nudging, you really shouldn’t have to do this; (this feature can be turned on or off in HD but not LE). If you are having to do this, it’s possible you have a larger (unusual) problem with your setup, record a test bus with some loopback patch cables, reverse the phase & test your system to be sure you don’t have a fault…

Cheers

N

[NickH]

Originally Posted by Johnkenn

I'm getting latency monitoring vocals when I sing...[Snip] I guess they only have ADC in the HD rigs to encourage you to buy the HD rig...

If you don’t turn off the delay compensation in HD when tracking, you get latency, (which is much more noticeable with the higher latency plug-in’s) just like you are getting with LE, so ADC is not the fix for this, it’s the TDM architecture that solves it.

N

[mookie]

i was asking some of the same questions on another thread... see http://gearslutz.com/board/showthread.php?t=100761

[Ronster]

Originally Posted by NickH

Pro Tools LE automatically compensates for any hardware delay after recording & does its own automatic nudging, you really shouldn’t have to do this; (this feature can be turned on or off in HD but not LE). If you are having to do this, it’s possible you have a larger (unusual) problem with your setup, record a test bus with some loopback patch cables, reverse the phase & test your system to be sure you don’t have a fault…

Cheers

N

actually, if you look up latency on the MBOX1 through google or DUC, you will see that there is no other way around the latency problem with the mbox1 other than "nudge"...other LE systems like the 001 or 002 don't have these problems...don't know if the mbox2 as the mbox1 because i haven't used that unit...only the mbox1.

[Johnkenn]

Well - I think I have whipped the mighty PT's latency. I did a quick test and this seemed to work.
I have a little Behringer mixer that cost me like - $40...So, I put it before the Apogee so I could monitor the signal in. I go out of my compressor into the line in of the mixer, go out of my mixer into the Apogee. Then spdif into the mbox and then spdif out of the mbox directly to the Apogee. I then go analog out to my Big Knob and one stereo output going to my monitors and another going back to the mixer so I can hear the mix. That way, there's no latency involved. It's actually kind of nice - I can control the two mix and the vocal levels...

[Ronster]

Originally Posted by Johnkenn

Well - I think I have whipped the mighty PT's latency. I did a quick test and this seemed to work.
I have a little Behringer mixer that cost me like - $40...So, I put it before the Apogee so I could monitor the signal in. I go out of my compressor into the line in of the mixer, go out of my mixer into the Apogee. Then spdif into the mbox and then spdif out of the mbox directly to the Apogee. I then go analog out to my Big Knob and one stereo output going to my monitors and another going back to the mixer so I can hear the mix. That way, there's no latency involved. It's actually kind of nice - I can control the two mix and the vocal levels...

cool, thanks for the update.

[NickH]

Originally Posted by Ronster

actually, if you look up latency on the MBOX1 through google or DUC, you will see that there is no other way around the latency problem with the mbox1 other than "nudge"...other LE systems like the 001 or 002 don't have these problems...don't know if the mbox2 as the mbox1 because i haven't used that unit...only the mbox1.

Actually that was a bug with the mbox that was fixed around Pro Tools Version 6.2…

So if you are still getting it with ver7.x, you have much bigger problems like I said before…

Originally Posted by Johnkenn

I have a little Behringer mixer that cost me like - $40...So, I put it before the Apogee so I could monitor the signal in. I go out of my compressor into the line in of the mixer, go out of my mixer into the Apogee...

Congrats upon finding a solution!

My only concern with this is that every signal you record is going through the line amps on that Behringer mixer, I am not sure if the pre’s & line inputs use the same pre circuit like they do on the ADA8000, but even so, maybe a splitting system or a higher quality mixer could be in order if you are recording something important…

Cheers

N

[Ronster]

Originally Posted by NickH

Actually that was a bug with the mbox that was fixed around Pro Tools Version 6.2…

So if you are still getting it with ver7.x, you have much bigger problems like I said before…

Congrats upon finding a solution!

My only concern with this is that every signal you record is going through the line amps on that Behringer mixer, I am not sure if the pre’s & line inputs use the same pre circuit like they do on the ADA8000, but even so, maybe a splitting system or a higher quality mixer could be in order if you are recording something important…

Cheers

N

ok, i get it know...i haven't used the mbox in a long time so maybe it's time to break out again and mess around with it for a few hours. i now have version 7 of Pt's, so maybe i'll enjoy working with it a lot more now.

and yeah, congrats on finding a way to make it work for you.

[theblue1]

Originally Posted by Ronster

i'm not sure about the mbox 2, but i've used the first mbox and i always had horrible latency issues with it. after doing some research on google and on the digi forum i found out 2 things:

1) mbox has a latency of something like 164 samples or something like that...after reading up on it i found out that i had to nudge each track back 164 samples after recording it

2) the best way to really avoid horrible latency that occurs in the monitoring was for me to mute the track that was in record mode while i was recording...i know it sounds strange but it worked.

don't know if any of this applies to the mbox2 but i just thought i'd throw these things out there in case it might help you...hopefully someone can chime in and confirm this. let me know if any of this helps with the mbox2!

Many DAW and hardware combinations leave the user with this sort of "overdub misalignment" -- unless they have a way of compensating. Mackie's Tracktion has a calibration-auto-track alignment utility. Cubase's hardware compensation appears to be somewhat similar. Sonar 6 has apparently added something to deal with it.

I get a 356 sample misalignment (about 8 ms at 44.1) using my MOTU 828mkII FW box. In Sonar 5, with no alignment comp, I have to nudge the tracks by that amount to get them to line up.

In Mackie Tracktion, I can set its track alignment comp on (after running a calibration diagnostic) and it handles it for me automatically. I think Cubase's ping loopback diagnostic is probably similar. (I don't know precisely how Sonar 6 dealt with the issue.)


I first became aware of this misalignment issue when reading about the original 001, 002, and Mbox on the DUC. All the interfaces had some amount of misalignment (although the PCI based 001 had a fairly negligible amount, something like 1.5 ms, IIRC).

I was feeling all superior until I did a ping loopback with my PCI-based Echo Mia and found about a 4.5 ms misalignment between each new track and previously recorded ones. I didn't feel so superior, after that.

Far as I know, all the PT LE systems still have this uncompensated misalignment -- but I could be wrong. Perhaps a comp utility made its way in in the last upgrade or two.

And almost everyone I've ever got to perform a loopback test -- on any DAW -- has discovered some amount of misalignment between new tracks and old -- unless they were using some form of alignment comp.

The issue blew into a big thread over at Harmony Central and Ron Kuper (MR. Cakewalk, to you) even came in to defend the program and say it was the drivers that were at fault.

But Kuper promised a fix when even Craig Anderton had a fairly nasty track misalignment -- and Kuper had to admit that he had a one sample misalignment -- and otherwise loyal Sonar users ganged up to say, Look, if almost EVERYONE has this issue to some extent, then clearly the system between hardware, drivers, and DAW is not working properly and the DAW-ware is a logical place to fix it -- particularly since other DAW programs DO offer some kind of track alignment compensation.


Anyhow, a lot of folks figure a few silly little milliseconds aren't that big a deal. Me, I'm still nudging until I get around to updating my software.

[Johnkenn]

Originally Posted by NickH

[i]

My only concern with this is that every signal you record is going through the line amps on that Behringer mixer, I am not sure if the pre’s & line inputs use the same pre circuit like they do on the ADA8000, but even so, maybe a splitting system or a higher quality mixer could be in order if you are recording something important…

Cheers

N

I spoke too soon. I'm going to have to get a mixer with some sort of monitoring section that allows you to connect something but not have it routed to the outputs of the mixer. Here's what's happening.
I'm going line in from the pre to the mixer and line out from the mixer to the apogee - then going spdif to the mbox. However, in order to monitor the mix, I'm having to go spdif out of the mbox to the apogee, then analog to the Big Knob where I have one out going to the monitors and another dedicated out back to the mixer in order to hear the tracks that I'm singing to. Problem is, when I go line in from the Big Knob to the mixer, the tracks are also being routed through the out of the mixer, so basically, I'm getting a vocal recorded with a low level track behind it.
This is a tiny 3 channel mixer with only one set of outs, so I think I'm going to have to get another mixer and put the BK first in the signal path and the mixer last. The BK has a monitor section that I could connect the 2 track mix to that wouldn't bleed into the vocals...Oy vey...How much trouble can this be?

[chrisrulesmore]

I have never worked with the Mbox 2, but I did have a Rosetta 200 hooked up to a Mbox 1. As best I can recall, I would get around the latency by muting the track I was recording (to mute the latent playback), whilst running the analog outs from your Rosetta into the Mbox along with the SPDIF signal. You still have the input in PT set to SPDIF, so that is what you are recording. The analog input from your Rosetta is only for direct monitoring.

This should work like a champ.

If you are looking to switch interfaces such that you have onboard direct monitoring, go with the M-Audio 1814 (or perhaps the 410, but I'm not sure) and PT M-Powered. This is way cheaper than the 002 which still only has low latency monitoring I believe.

-Chris

[Ronster]

Originally Posted by theblue1

Many DAW and hardware combinations leave the user with this sort of "overdub misalignment" -- unless they have a way of compensating. Mackie's Tracktion has a calibration-auto-track alignment utility. Cubase's hardware compensation appears to be somewhat similar. Sonar 6 has apparently added something to deal with it.

I get a 356 sample misalignment (about 8 ms at 44.1) using my MOTU 828mkII FW box. In Sonar 5, with no alignment comp, I have to nudge the tracks by that amount to get them to line up.

In Mackie Tracktion, I can set its track alignment comp on (after running a calibration diagnostic) and it handles it for me automatically. I think Cubase's ping loopback diagnostic is probably similar. (I don't know precisely how Sonar 6 dealt with the issue.)


I first became aware of this misalignment issue when reading about the original 001, 002, and Mbox on the DUC. All the interfaces had some amount of misalignment (although the PCI based 001 had a fairly negligible amount, something like 1.5 ms, IIRC).

I was feeling all superior until I did a ping loopback with my PCI-based Echo Mia and found about a 4.5 ms misalignment between each new track and previously recorded ones. I didn't feel so superior, after that.

Far as I know, all the PT LE systems still have this uncompensated misalignment -- but I could be wrong. Perhaps a comp utility made its way in in the last upgrade or two.

And almost everyone I've ever got to perform a loopback test -- on any DAW -- has discovered some amount of misalignment between new tracks and old -- unless they were using some form of alignment comp.

The issue blew into a big thread over at Harmony Central and Ron Kuper (MR. Cakewalk, to you) even came in to defend the program and say it was the drivers that were at fault.

But Kuper promised a fix when even Craig Anderton had a fairly nasty track misalignment -- and Kuper had to admit that he had a one sample misalignment -- and otherwise loyal Sonar users ganged up to say, Look, if almost EVERYONE has this issue to some extent, then clearly the system between hardware, drivers, and DAW is not working properly and the DAW-ware is a logical place to fix it -- particularly since other DAW programs DO offer some kind of track alignment compensation.


Anyhow, a lot of folks figure a few silly little milliseconds aren't that big a deal. Me, I'm still nudging until I get around to updating my software.

thanks for posting this...i've been looking for some information on latency with the digi001 but i couldn't really find any. so there is latency with the 001 huh? i guess i'm just going to have to perform a latency test and see what kind of latency i'm getting...so that i can fix it with the nudge feature. i noticed you suggested the loop back test, i've heard of people doing this but have never tried it myself...i guess it's time to give it a try. i'm not to sure about Pro Tools having fixed uncompensated misalignment in the later versions, i'm gonna have to look into it...but i'm pretty sure it's still there in LE systems running the most recent versions.

[theblue1]

Well, it gets confusing because there are so many different latencies that come up. And they're all somewhat related but quite different, as well. There's the converters internal latency and then the converter's i/o buffers. There are buffers in your software to accomodate for various computer-side latencies, DAW processing, plugs, etc. We typically see that in the form of monitoring buffer latency settings in our DAW/recording software.

And then there is the peculiar issue of track misalignment, clearly related to uncompensated or improperly compensated hardware and or software processing latencies.

Obviously, in a perfect world, our converter's drivers would report a proper latency setting in sample count to the DAW which would properly adjust newly recorded tracks by the proper amount to get them to align precisely with previously recorded tracks.

(As we tape-generation types had always come to expect.)

But, in the real world, it appears many (maybe even most) of us have some sort of misalignment between new tracks and previously recorded tracks.


It sounds like you know what the issues are but let me just go through the overview of loopback testing...

You'll use a previously recorded sound with a sharp, easily identifiable transient. You could generate a low amplitude square wave, if you wanted and use that. You just want to be able to get your bearing to the precise sample number, if at all possible. We'll call this the "ping." It can be short duration but not so short you can't find it in your audio editor.

Turn your control room monitors all the way down, just to be on the safe side.

Now route your ping out an analog output and route that analog out into one of your analog ins. Do NOT source monitor that input which could create a feedback loop from hell.

Now simply record the ping playback on a new track.

It sounds like it's obvious to you why these SHOULD line up to the sample.

Sometimes folks have trouble understanding the motivation for a loopback test: as simply as I can put it, you want what's playing NOW through your monitoring to line up with what you're recording NOW. A loopback test is a way of measuring any gap in alignment.

Once you've sent your ping out analog and recorded it back in, zoom into the wave view of it in your DAW. Zoom into the original ping and note the precise sample address of the leading edge of the transient. Do the same with the copy.

The copy is likely behind the original but might be in front of it. Subtract to find the difference between them and and then set a nudge to that precise amount.

In my DAW, if a clip begins at 0:00, I can't nudge it backwards any farther. Which means I have to slip edit the clip away from zero and then nudge. At first that irritated me but then I realized I could use that as a way to know whether or not I'd nudged a clip... if it still started flush at 0, I hadn't nudged it. Kludgy, I know, but if you got a lemon tree, make lemonade.

Anyhow, if there's NO difference between your original ping and the copy, you win, your rig is already, shall we say, time-aligned. No, we shan't. But anyhow, you don't have to do anything, in that case. It's working as it should.

[Johnkenn]

Originally Posted by chrisrulesmore

I have never worked with the Mbox 2, but I did have a Rosetta 200 hooked up to a Mbox 1. As best I can recall, I would get around the latency by muting the track I was recording (to mute the latent playback), whilst running the analog outs from your Rosetta into the Mbox along with the SPDIF signal. You still have the input in PT set to SPDIF, so that is what you are recording. The analog input from your Rosetta is only for direct monitoring.

This should work like a champ.

If you are looking to switch interfaces such that you have onboard direct monitoring, go with the M-Audio 1814 (or perhaps the 410, but I'm not sure) and PT M-Powered. This is way cheaper than the 002 which still only has low latency monitoring I believe.

-Chris

Damn - wish I had known this before I went with the Mbox...

[Jose Mrochek]

umm I recently bought an mbox 2pro with a macbook, and the latency seems fine, but I'm kinda slow.. so maybe my brain's latency syncs perfectly ?? haha

[Jim Kerr]

I think using a mixer as your analog interface is always the way to go. Yes the mbox1 has input monitoring and it is a blend of input and output. So you mute the track your recording on to and monitor your input while its blended with the output, turn up, done. I think the post above (Rosetta analog out > in mbox while spdf is the ((real)) input solves the problem.

To be quite honest all of this is quite normal. If you were going to tape, real tape you would not want to monitor your playback of the track you were recording on either because there would be...drum roll......latency. In form of the distance of the record head and playback head....right?!? Thats why there are consoles with all the really great fun features.